Merge branch 'net-allow-virtual-netdevs-to-forward-udp-l4-and-fraglist-gso-skbs'

Alexander Lobakin says:

====================
net: allow virtual netdevs to forward UDP L4 and fraglist GSO skbs

NETIF_F_GSO_UDP_L4 and NETIF_F_GSO_FRAGLIST allow drivers to offload
GSO UDP L4. This works well on simple setups, but when any logical
netdev (e.g. VLAN) is present, kernel stack always performs software
resegmentation which actually kills the performance.

The full path in such cases is like:
1. Our NIC driver advertises a support for fraglists, GSO UDP L4, GSO
   fraglists.
2. User enables fraglisted GRO via Ethtool.
3. GRO subsystem receives UDP frames from driver and merges the packets
   into fraglisted GSO skb(s).
4. Networking stack queues it up for xmitting.
5. Virtual device like VLAN doesn't advertise a support for GSO UDP L4
   and GSO fraglists, so skb_gso_check() doesn't allow to pass this skb
   as is to the real driver.
6. Kernel then has to form a bunch of regular UDP skbs from that one and
   pass it to the driver instead. This fallback is *extremely* slow for
   any GSO types, but especially for GSO fraglists.
7. All further processing performs with a series of plain UDP skbs, and
   the driver gets it one-by-one, despite that it supports UDP L4 and
   fraglisted GSO.

That's not OK because:
a) logical/virtual netdevs like VLANs, bridges etc. should pass GSO skbs
   as is;
b) even if the final driver doesn't support such type of GSO, this
   software resegmenting should be performed right before it, not in the
   middle of processing -- I think I even saw that note somewhere in
   kernel documentation, and it's totally reasonable in terms of
   performance.

Despite the fact that no mainline drivers currently supports fraglist
GSO, this should and can be easily fixed by adding UDP L4 and fraglist
GSO to the list of GSO types that can be passed-through the logical
interfaces (NETIF_F_GSO_SOFTWARE). After this change, no resegmentation
occurs (if a particular driver supports and advertises this), and the
performance goes on par with e.g. 1:1 forwarding.
The only logical netdevs that seem to be unaffected to this are bridge
interfaces, as their code uses full NETIF_F_GSO_MASK.

Tested on MIPS32 R2 router board with a WIP NIC driver in VLAN NAT:
20 Mbps baseline, 1 Gbps / link speed with this patch.

Since v1 [1]:
 - handle bonding and team drivers as suggested by Willem de Bruijn;
 - reword and expand the introduction with the particular example.

[1] https://lore.kernel.org/netdev/Mx3BWGop6fGORN6Cpo4mHIHz2b1bb0eLxeMG8vsijnk@cp3-web-020.plabs.ch
====================

Link: https://lore.kernel.org/r/NysZRGMkuWq0KPTCJ1Dz2FTjRkeJXDH3edVrsEeJkQI@cp4-web-036.plabs.ch
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

5 files changed, 13 insertions, 16 deletions

diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 84ecbc6fa0ff..71c9677d135f 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -1228,14 +1228,14 @@ static netdev_features_t bond_fix_features(struct net_device *dev,
 }
 
 #define BOND_VLAN_FEATURES	(NETIF_F_HW_CSUM | NETIF_F_SG | \
-				 NETIF_F_FRAGLIST | NETIF_F_ALL_TSO | \
+				 NETIF_F_FRAGLIST | NETIF_F_GSO_SOFTWARE | \
 				 NETIF_F_HIGHDMA | NETIF_F_LRO)
 
 #define BOND_ENC_FEATURES	(NETIF_F_HW_CSUM | NETIF_F_SG | \
-				 NETIF_F_RXCSUM | NETIF_F_ALL_TSO)
+				 NETIF_F_RXCSUM | NETIF_F_GSO_SOFTWARE)
 
 #define BOND_MPLS_FEATURES	(NETIF_F_HW_CSUM | NETIF_F_SG | \
-				 NETIF_F_ALL_TSO)
+				 NETIF_F_GSO_SOFTWARE)
 
 
 static void bond_compute_features(struct bonding *bond)
@@ -1291,8 +1291,7 @@ done:
 	bond_dev->vlan_features = vlan_features;
 	bond_dev->hw_enc_features = enc_features | NETIF_F_GSO_ENCAP_ALL |
 				    NETIF_F_HW_VLAN_CTAG_TX |
-				    NETIF_F_HW_VLAN_STAG_TX |
-				    NETIF_F_GSO_UDP_L4;
+				    NETIF_F_HW_VLAN_STAG_TX;
 #ifdef CONFIG_XFRM_OFFLOAD
 	bond_dev->hw_enc_features |= xfrm_features;
 #endif /* CONFIG_XFRM_OFFLOAD */
@@ -4721,7 +4720,7 @@ void bond_setup(struct net_device *bond_dev)
 				NETIF_F_HW_VLAN_CTAG_RX |
 				NETIF_F_HW_VLAN_CTAG_FILTER;
 
-	bond_dev->hw_features |= NETIF_F_GSO_ENCAP_ALL | NETIF_F_GSO_UDP_L4;
+	bond_dev->hw_features |= NETIF_F_GSO_ENCAP_ALL;
 #ifdef CONFIG_XFRM_OFFLOAD
 	bond_dev->hw_features |= BOND_XFRM_FEATURES;
 #endif /* CONFIG_XFRM_OFFLOAD */
diff --git a/drivers/net/dummy.c b/drivers/net/dummy.c
index bab3a9bb5e6f..f82ad7419508 100644
--- a/drivers/net/dummy.c
+++ b/drivers/net/dummy.c
@@ -124,7 +124,7 @@ static void dummy_setup(struct net_device *dev)
 	dev->flags &= ~IFF_MULTICAST;
 	dev->priv_flags |= IFF_LIVE_ADDR_CHANGE | IFF_NO_QUEUE;
 	dev->features	|= NETIF_F_SG | NETIF_F_FRAGLIST;
-	dev->features	|= NETIF_F_ALL_TSO;
+	dev->features	|= NETIF_F_GSO_SOFTWARE;
 	dev->features	|= NETIF_F_HW_CSUM | NETIF_F_HIGHDMA | NETIF_F_LLTX;
 	dev->features	|= NETIF_F_GSO_ENCAP_ALL;
 	dev->hw_features |= dev->features;
diff --git a/drivers/net/ifb.c b/drivers/net/ifb.c
index 7fe306e76281..fa63d4dee0ba 100644
--- a/drivers/net/ifb.c
+++ b/drivers/net/ifb.c
@@ -187,8 +187,7 @@ static const struct net_device_ops ifb_netdev_ops = {
 };
 
 #define IFB_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG  | NETIF_F_FRAGLIST	| \
-		      NETIF_F_TSO_ECN | NETIF_F_TSO | NETIF_F_TSO6	| \
-		      NETIF_F_GSO_ENCAP_ALL 				| \
+		      NETIF_F_GSO_SOFTWARE | NETIF_F_GSO_ENCAP_ALL	| \
 		      NETIF_F_HIGHDMA | NETIF_F_HW_VLAN_CTAG_TX		| \
 		      NETIF_F_HW_VLAN_STAG_TX)
 
diff --git a/drivers/net/team/team.c b/drivers/net/team/team.c
index 07f1f3933927..b4092127a92c 100644
--- a/drivers/net/team/team.c
+++ b/drivers/net/team/team.c
@@ -975,11 +975,11 @@ static void team_port_disable(struct team *team,
 }
 
 #define TEAM_VLAN_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \
-			    NETIF_F_FRAGLIST | NETIF_F_ALL_TSO | \
+			    NETIF_F_FRAGLIST | NETIF_F_GSO_SOFTWARE | \
 			    NETIF_F_HIGHDMA | NETIF_F_LRO)
 
 #define TEAM_ENC_FEATURES	(NETIF_F_HW_CSUM | NETIF_F_SG | \
-				 NETIF_F_RXCSUM | NETIF_F_ALL_TSO)
+				 NETIF_F_RXCSUM | NETIF_F_GSO_SOFTWARE)
 
 static void __team_compute_features(struct team *team)
 {
@@ -1009,8 +1009,7 @@ static void __team_compute_features(struct team *team)
 	team->dev->vlan_features = vlan_features;
 	team->dev->hw_enc_features = enc_features | NETIF_F_GSO_ENCAP_ALL |
 				     NETIF_F_HW_VLAN_CTAG_TX |
-				     NETIF_F_HW_VLAN_STAG_TX |
-				     NETIF_F_GSO_UDP_L4;
+				     NETIF_F_HW_VLAN_STAG_TX;
 	team->dev->hard_header_len = max_hard_header_len;
 
 	team->dev->priv_flags &= ~IFF_XMIT_DST_RELEASE;
@@ -2175,7 +2174,7 @@ static void team_setup(struct net_device *dev)
 			   NETIF_F_HW_VLAN_CTAG_RX |
 			   NETIF_F_HW_VLAN_CTAG_FILTER;
 
-	dev->hw_features |= NETIF_F_GSO_ENCAP_ALL | NETIF_F_GSO_UDP_L4;
+	dev->hw_features |= NETIF_F_GSO_ENCAP_ALL;
 	dev->features |= dev->hw_features;
 	dev->features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_STAG_TX;
 }
diff --git a/include/linux/netdev_features.h b/include/linux/netdev_features.h
index 0b17c4322b09..934de56644e7 100644
--- a/include/linux/netdev_features.h
+++ b/include/linux/netdev_features.h
@@ -207,8 +207,8 @@ static inline int find_next_netdev_feature(u64 feature, unsigned long start)
 				 NETIF_F_FSO)
 
 /* List of features with software fallbacks. */
-#define NETIF_F_GSO_SOFTWARE	(NETIF_F_ALL_TSO | \
-				 NETIF_F_GSO_SCTP)
+#define NETIF_F_GSO_SOFTWARE	(NETIF_F_ALL_TSO | NETIF_F_GSO_SCTP |	     \
+				 NETIF_F_GSO_UDP_L4 | NETIF_F_GSO_FRAGLIST)
 
 /*
  * If one device supports one of these features, then enable them